Our initial (1985) neuropathologic/virologic studies on AIDS encephalopathy in children identified the brain as a primary site of persistent HIV-1 infection. The brain is a reservoir for large amounts of HIV-1 unintegrated DNA (UVD) and is the major site of multinucleated (syncytial) cell formation. In vitro passage of HIV-1 selects for replication efficiency in the target cells employed, but excludes HIV-1 strain variants important for neuropathology. In particular, most laboratory strains of HIV-1 are lymphocyte-tropic, whereas much recent evidence points to a link between macrophage-tropic strains and neurovirulence. Studies from our groups have implicated neuroinvasion or neurovirulence in AIDS with specific regions of the HIV-1 env gp120 (CD4 binding, V3 and "efficiency" domains), and with the regulatory genes nef and vif. Our Specific Aims are: 1. To amplify and clone these HIV-1 gene sequences directly from post-mortem brain and selected tissues of children. 2. To determine the nucleotide sequence of these HIV-1 genes and characterize differences between individuals, between different tissues of the same individual (including PBMC), and between tissue-derived and cultured HIV- 1 isolates. 3. To construct chimeric HIV-1 DNAs using lymphocyte- trophic, macrophage-trophic, or glial-trophic HIV-1 clones with selected neuropathologic gene sequences. These recombinant constructs will be used to test HIV-1 replication, syncytium formation and cytokine induction in neural tissue (Project 1), and neurotoxicity (Projects 2 and 3) in both in vitro and in vivo assay systems.